A real-time model of the cerebellar circuitry underlying classical conditioning: A combined simulation and robotics study

Although key components of the cerebellar circuitry relevant to classical conditioning have been identified, the question how they act together is still unresolved. In this simulation study, we investigate a real-time model which captures basic anatomical and physiological properties of this system. We show that this model displays realistic learning performance over a range of inter-stimulus intervals, and demonstrate its stability using a mobile robot solving an obstacle avoidance task.